Simplify inference to one sequence#12
Merged
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Match CotabbyInference to the shipping product workload: one model, one context, and one live autocomplete sequence. Remove the unused second slot, decoder worker/queue/batching machinery, dead chat/manual-logit/snapshot/test-only APIs, the ineffective native token-budget field, and the obsolete Phase-0 benchmark. Refresh the public README and repository guidance, and ignore local
.internal/maintainer notes.The engine still uses a changing external sequence ID over fixed llama.cpp slot zero, preserving stale-handle rejection after resets while cutting unused context capacity and native lifecycle states.
Validation
swift test— 14 tests, 7 model-backed skips, 0 failures.COTABBY_TEST_MODEL_PATH=.../Qwen3.5-0.8B-Base.i1-Q6_K.gguf swift test— 14 tests, 0 failures.COTABBY_TEST_MODEL_PATH=.../SmolLM2-135M-Instruct-q8_0.gguf swift test— 14 tests, 0 failures.xcodebuild ... buildagainst package commit8f3979d— exit 0.git diff --check— clean.Risk / rollout notes
This intentionally removes unused public API. The only current Cotabby caller affected by the aggregate layout was updated and merged before this PR. Partial KV trim remains model-capability-dependent; Cotabby already rebuilds the sequence when hybrid/SWA caches reject it.